The design of a mobile concentrator plant for gold extraction form tailings in Tanzania

Abstract: The advantages brought by a mobile concentrator plant are many.
Nevertheless, the knowledge of the materials to be processed is of great
importance for the process design, which in turn is used for plant design.
Tailings samples from five different locations, which include Tanga,
Morogoro, Dodoma, Mawemeru and Mgombani in Mwanza region, were
characterized by x-ray diffraction (XRD), optical microscopy, dry sieve and
cyanide leaching analyses. Subsequently, the characterization results were
used to design the process route for gold extraction from tailings and
equipment selection of a mobile plant. Using MODSIM digital simulator and a
simple model developed for prediction of free gold recovery, mass balances
and optimal process were obtained, which form a base for the economic
evaluation of the process.

It was found that gold occurs as free grains of 5 – 20 µm size within both
fine and coarse fractions of the tailings deposits at a magnification size
of 200 – 500x. It was noted that leaching by cyanide reaches its
equilibrium and optimum gold dissolution after 23 hours. The recoveries of
gold were 80 % and 91 % for composite and Tanga tailings deposits,
respectively. Recovery of 91 % gold by cyanide leaching confirmed the
presence of free gold observed by optical microscopy. Gold was also
associated with silicate minerals, more especially quartz and feldspars and
in some cases with oxides such as goethite. Other minerals found in the
tailings deposits were chalcopyrite, pyrite, pyrrhotite and hematite for
Tanga, Dodoma, Mawemeru and Mgombani, whereas silver telluride and chromium
telluride showed a unique character for Morogoro tailings sample.
The presence of chalcopyrite and silver minerals showed potentiality of
copper and silver production in the gold concentrate. An uncertain
character of gold smeared onto other mineral grains was noted, but the
exact composition has to be verified by electron microprobe or scanning
electron microscopic analyses.

Moreover, particle size analysis showed that only Mgombani tailings (24.61
g/t Au) contains high percentage of fairly fine particle fraction whose P.
80 % is 138 µm with high gold concentration in the fine fraction (< 45 µm)
contrary to dodoma tailings (0.54 g/t au) which showed p. 80 % of the
material is 586 µm with high gold concentration in coarse fraction. the
rest of the tailings deposits have shown that p. 80 % of the materials
ranges between 230 – 377 µm with corresponding feed grade 5.4 g/t au, 3.8
g/t au and 3.41 g/t au for tanga, morogoro and mawemeru with normal gold
distribution, respectively. uncertainty of the presence of free fine gold
in the coarse fractions was due to dry sieving analysis. these suggested
that configuration of the drum washer, vibrating screen, knelson
concentrator and two stages gravity concentration by primary spiral
concentrator and shaking table cleaning concentrator is a suitable process
for gold separation from all the deposits irrespective of their variations.

simulation technique has been used to analyze the proposed process. a
simple model for prediction of the recovery of free gold from tailings by
knelson concentrator was developed and used. it must be noted that the
model is valid at 0.8 % minimum allowable gangue recovery into concentrate
for particle size < 678 µm and 1.2% maximum allowable gangue recovery for
particle size > 678 µm. The model predicted the optimum concentrate grade
ranges from 93 – 5007 g/t for materials with lower feed grade (0.54 g/t)
and highest grade (24.6 g/t), respectively, at the cut size of 160 µm. A
feed grade of 3.41 g/t Au gives a concentrate grade of 510 g/t Au and was
used to evaluate the economic viability of the designed process. Low
concentrate grade from the configuration of the spiral and shaking table is
due to inability of the MODSIM models to predict the performance of the
spiral concentrator for gold deposits.

The proposed mobile concentrator plant of capacity 5 t/h is economically
viable with a payback period of one year. The economic evaluation shows
that NPV at 10 % interest rate is estimated to be $ 26,323,483 which is
much higher for investing the same amount of money in the bank ( NPV $
296,403). It means that it is better for an investor to invest in the
designed plant for gold extraction from tailings rather than to deposit
money for the bank investment. The economic analysis identified that the
most severe risk parameters for the designed process are concentrate grade,
gold price and operating costs. Sensitivity analysis showed that the
minimum concentrate grade should be 155 g/t for investing on a designed
plant. A pilot study is suggested in order to increase the knowledge of the
proposed plant and resources.